Thymic Epithelial Cell Support of Thymopoiesis Does Not Require Klotho

Thymic aging is a major contributing factor to immunological senescence as the thymus begins to involute with the onset of puberty, leading to a progressive defect in the ability to generate naive T cells (1, 2). The primary elements of the thymic microenvironment affected by age-related involution are structural thymic epithelial cells (TECs). TECs are replaced by adipocytes and peripheral lymphocytes, resulting in decreased thymopoietic activity and T cell selection (3). Consequently, both quantity and quality of the T cell repertoire are affected with decreased de novo T cell generation, which is compensated by a homeostatic oligoclonal expansion of T cells in the periphery (4, 5). These dynamic changes result in an increased susceptibility to infections (6, 7), suboptimal responses to vaccines (8–12), and an increased risk to develop cancer and autoimmune diseases (13–15). The age-related decline in thymopoietic activity (1) is especially apparent in patients who have undergone chemotherapy (16) or allogeneic hematopoietic stem cell (HSC) transplantation (17). The necessary preparative regimen with cytotoxic chemotherapy and/or radiation severely damages the thymus, the recovery of which is extremely limited in aged individuals (18, 19).

To study the process of aging in mice, Klotho-deficient (Kl/Kl) animals have been used, as they grow normally up to 3 wk of age and then begin to show premature aging phenotypes (20). Klotho encodes a β-glucuronidase–related molecule in two separate isoforms, transmembrane and secreted; the transmembrane molecule serves as a coreceptor for fibroblast growth factor 23 (FGF23) by transporting this cytokine to its receptor, FGFR1c, and thereby regulating mineral metabolism (21–23). Klotho is expressed in the kidney and parathyroid gland, and the secreted form can also be found in the blood, cerebrospinal fluid, and urine (24). FGF23 suppresses phosphate reabsorption and vitamin D synthesis in the kidney, causing negative phosphate balance because of both its phosphaturic hormone function and its function as a counter-regulatory hormone for vitamin D (24). The secreted form of Klotho inhibits insulin growth factor 1 signaling and confers increased resistance to oxidative stress (25–27). Mice transgenic for Klotho live 20–30% longer than wild-type (WT) controls (28), whereas the protein’s absence results in an advanced aging syndrome resembling progeria. Multiple organs are affected in Kl/Kl mice, resulting in growth retardation, pituitary abnormalities, arteriosclerosis, ectopic calcification of various organs, osteoporosis, skin atrophy, emphysema, and atrophy of both the genital organs and the thymus (20). Interestingly, mice that are FGF23 deficient or Kl/Kl have phenotypes similar to one another. These deficits can be alleviated by reversing the effects of hyperphosphatemia either genetically or by diet, suggesting a link between aging and phosphate (24).

The Kl/Kl mouse model has provided insight into the process of aging in humans. Indeed, human KLOTHO shares 86% amino acid identity with its mouse ortholog (29). Individuals homozygous for KLOTHO variants that disrupt the molecule’s trafficking and catalytic functions experience a decreased life expectancy (29); have increased cardiovascular risk factors, such as elevated high-density lipoprotein cholesterol levels and high systolic blood pressure (30); and demonstrate an increased risk for stroke and coronary artery disease (31). Polymorphisms in KLOTHO (loss of function) have been associated with an increased risk for osteoporosis and spondylosis (32), and reduced KLOTHO protein expression has been noted in patients with chronic renal failure (33).

Although the effects of Klotho deficiency on kidney development and function, mineral metabolism, and bone maintenance are well studied in Kl/Kl mice, the direct effect of Klotho deficiency on the TECs is unknown. We therefore sought to determine whether the effects of Klotho on thymic aging are cell intrinsic or reflect a systemic metabolic consequence of a lack of the Klotho protein.

B6.Cg-Foxn1^nu/J mice were purchased from The Jackson Laboratory and were used at 8–12 wk of age. Kl/+ mice (B6-CD45.2⁺) were generously provided by the University of California, Davis, Mutant Mouse Regional Resource Centers and were intercrossed (Kl/+ by Kl/+) in our animal colony under the guidance of in-house veterinary staff. B6-Ly5.2/Cr (B6-CD45.1⁺) mice were purchased from the National Cancer Institute and were used at 7 wk of age. Mice were housed in a specific pathogen-free facility and used with the approval of the University of Minnesota Institutional Animal Care and Use Committee. For vitamin D experiments, breeders were fed and pups were maintained on a vitamin D–deprived diet (TD 89123) purchased from Harlan Teklad.

Mouse thymus, spleen, and lymph node were processed into single-cell suspensions and analyzed by flow cytometry. TECs were isolated as previously described (34). Dead cells were stained by a fixable viability dye conjugated to eFluor 780 (eBioscience). Fixation and intracellular/intranuclear staining were performed using the eBioscience Foxp3 staining kit or BD Fixation/Permeabilization Solution Kit. The following Abs were purchased from BD Biosciences: TCRβ (H57-597), TCRγδ (GL-3), and EpCAM (G8.8). The following Abs were purchased from eBioscience: CD3 (145-2C11), CD4 (GK1.5), CD11c (N418), CD25 (PC61.5), CD45R (B220, RA3-6B2), MHC class II (M5/114), and Ki67 (SolA15). The following Abs were purchased from BioLegend: CD8 (53-6.7), CD11b (M1/70), CD44 (IM7), CD45 (30-F11), CD62L (MEL-14), CD122 (TM-β1), Gr-1 (RB6-8C5), Ly-51 (6C3), NK-1.1 (PK136), and TER-119 (TER-119). UEA-1 was purchased from Vector Laboratories. Data were collected using a BD LSR II flow cytometer and were analyzed using FlowJo version 10 (Tree Star). The TaqMan Gene Expression Assay (assay identifiction number: Mm00502000_m1; Thermo Fisher Scientific) was used for quantification of Klotho gene expression.

Mice heterozygous for Klotho were mated overnight and then separated. At the time of harvest, neonate pups were screened for Klotho via PCR. WT or Kl/Kl thymi were placed under the kidney capsule of B6.Cg-Foxn1^nu/J mice in the previously described manner (35, 36).

B6-CD45.1⁺ recipients were lethally irradiated using 1100 cGy total body irradiation by x-ray 1 d before infusion. On the second day, bone marrow cells (BM) were harvested from Kl/Kl mice and littermates. Mature T cells were removed from donor BM using anti-CD4 and anti-CD8 Abs, and low-toxicity rabbit complement was given i.v. at a cell dose of 1 × 10⁷.

Thymi were harvested and snap frozen in O.C.T. compound. Frozen sections (8 μm) were cut using a CM1900 cryostat (Leica Biosystems). Slides were dried for 30 min and then were immerged in acetone for 5 min at room temperature. The sections were blocked in PBS with 3% BSA (PBSB) for 1 h at room temperature and stained with the rabbit anti-mouse K5 polyclonal Ab (MBL International) and rat anti-mouse K8 mAb (TROMA-I; Developmental Studies Hybridoma Bank, University of Iowa), followed by DyLight 550 donkey anti-rabbit IgG Ab and DyLight 650 donkey anti-rat IgG (Invitrogen). ProLong Gold antifade reagent (Invitrogen) was used to prevent photobleaching. Images were obtained using a microscope (DM5500B; Leica Biosystems) with a camera (DFC340FX; Leica Microsystems) operating with the Leica Application Suite Advanced Fluorescence (Leica Microsystems) software and analyzed using ImageJ (National Institutes of Health) software.

Prism software (GraphPad) was used for statistical analysis. Data sets were compared using an unpaired Mann–Whitney U test. Data are shown as mean values ± SD. Significance was defined as p < 0.05.

The thymus is the primary lymphoid organ for the development of T cells by providing a unique microenvironment able to attract T cell precursors from the blood and control their T cell lineage commitment, differentiation, and selection (37). The earliest stages of intrathymic T cell development are marked by the absence of CD4 and CD8 expression as well as other lineage markers (designated double-negative [DN] cells). These immature DN cells are located in the cortex, where they significantly expand in number before they acquire the concomitant expression of both CD4 and CD8 (double-positive [DP]) and express an αβ T cell Ag receptor (TCR). DP thymocytes constitute the most abundant subpopulation of thymocytes and are subjected first to a process of positive selection, which assures that thymocytes express a TCR with sufficient affinity for a peptide–MHC complex expressed on cortical TECs (cTECs) (38). Positively selected thymocytes differentiate into TCRβ⁺ CD4 and CD8 single-positive (SP) cells, respectively, and migrate to the medulla, continuing their selection and maturation (39, 40). Self-reactive thymocytes are either eliminated by negative selection (clonal deletion) or differentiate into regulatory T cells (Tregs) (41). Subsequently, thymocytes are selected for tolerance to tissue-restricted Ags but are responsive to foreign Ags presented by the individual’s MHC haplotype (39).

We investigated the thymic microenvironment of Kl/Kl mice from 4 wk of age, shortly after which these mice begin to display signs of advanced aging (20). At 4 wk of age, the thymi of Kl/Kl mice were comparable to those of their WT littermates in size, cellularity, and subpopulation distribution (Fig. 1A–C). Kl/Kl mice at 8 wk of age (young adult) displayed a significantly reduced thymic size when compared with that of WT littermate controls (Fig. 1A) and a profound decrease of thymocytes in total cellularity (Fig. 1B). Flow cytometry revealed that the frequency of DP thymocytes was significantly decreased but that of DN, CD4SP, and CD8SP thymocytes remained unchanged in 8-wk-old Kl/Kl mice when compared with age-matched controls (Fig. 1C, Supplemental Fig. 1A–D). All thymocyte subpopulations had, however, a significantly reduced cellularity in 8-wk-old Kl/Kl mice (Fig. 1D). Semimature SP thymocytes were barely detected in 8-wk-old Kl/Kl mice (Fig. 1E, 1F, Supplemental Fig. 1E), suggesting that de novo generation of SP thymocytes was severely blocked (42). Thus, the thymus of 8-wk-old Kl/Kl mice was prematurely involuted to an extent only observed in older, physiologically aged mice.

To assess the effect of Klotho on TECs, both cTECs and medullary TECs (mTECs) were evaluated. They are defined by Ly51 expression and binding capacity to the lectin UEA-1. cTECs are critical for T cell progenitor expansion and positive selection of thymocytes, whereas both cTECs and mTECs affect the negative selection of self-reactive thymocytes (43). To address whether TEC development and maintenance are impaired in the absence of Klotho, we assessed TEC cellularity and frequencies in 4- and 8-wk-old Kl/Kl mice. Total TEC cellularity and the frequencies of their individual subsets were comparable at 4 wk of age to those of WT or heterozygous littermate controls (Fig. 2A, 2B). In contrast, 8-wk-old Kl/Kl mice showed a significant reduction in the number of total TECs (Fig. 2A), affecting especially mTECs (Fig. 2B, 2C). These data suggest that a Klotho deficiency intrinsically or extrinsically reduced TEC cellularity in 8-wk-old animals.

To determine the expression of Klotho in TECs, we performed RT-PCR on sorted cTEC, mTEC^lo, and mTEC^hi cells in both WT and Kl/Kl animals. Klotho expression was found in both mTEC compartments but was minimal in the cTEC compartment in WT animals. As expected, there was no expression in the TECs of Kl/Kl animals. Thymic involution is the result of cell-autonomous changes in cell maintenance as a function of age but can also occur as a result of systemic abnormalities via various paracrine effector mechanisms. To distinguish between these two explanations for the observed presenescent changes in thymic cellularity, we grafted thymic lobes from either neonatal Kl/Kl or WT mice under the kidney capsule of haploidentical athymic [i.e., nude, Foxn1^nu/nu (44)] recipients and compared their growth and differentiation. Foxn1^nu/nu mice are homozygously deficient for the expression of functional Foxn1, a master regulator of TEC growth, differentiation, and function of TEC, and therefore athymic and T cell deficient, but they have otherwise intact HSCs. To have a chronotypic comparison with 8-wk-old Kl/Kl mice, thymus grafts and peripheral lymphoid tissues of grafted recipients were analyzed 8 wk after transplantation. The size and cellularity of Kl/Kl and WT thymus grafts were comparable (Fig. 3A, 3B) and displayed identical proportions and cellularity of the distinct thymocyte subpopulations (Fig. 3C). The ostensibly normal thymopoietic activity of Kl/Kl grafts resulted in a comparable peripheral T cell reconstitution in both groups of transplanted mice (Fig. 3E, 3F). In aggregate, these results demonstrated that non-hematopoietic thymic stromal cells, including TECs, do not rely on Klotho expression for the organs’ thymopoietic activity.

Although mature blood cells, including T cells, do not express Klotho, findings from the transplant experiments did not formally exclude the requirement of Klotho for progenitor cells to commit to a T cell fate and differentiate into mature thymocytes able to promote TEC maturation via thymus cross-talk. To examine this possibility, lethally irradiated H-2–matched WT mice were grafted with either Klotho-deficient or -proficient HSCs and analyzed 8 wk later (Fig. 4A). Recipients rescued with Kl/Kl HSCs displayed thymocyte cellularity and differentiation comparable to mice grafted with WT HSCs (Fig. 4B–D). These results demonstrate that thymic Klotho expression in thymocytes is dispensable for repopulation of the T cell lineage in irradiated hosts.

Previous studies have shown that elimination of vitamin D in Kl/Kl mice diet corrected the hypervitaminosis D typically observed in these animals and consequently improved some of their disease-related phenotype (25, 45, 46). As our transplantation data indicated that premature thymic involution in Kl/Kl mice was not caused by Klotho deficiency in either hematopoietic cells or thymic stromal cells, we next investigated whether a reduction in vitamin D levels improved the thymic changes and, as a result, the peripheral T cell compartment of Kl/Kl mice. In this experiment, both breeder mice and offspring were maintained on a vitamin D–deprived diet throughout the experiment. Eight-week-old Kl/Kl mice exposed since conception to low vitamin D levels displayed a total thymus cellularity and thymocyte differentiation comparable to those of age-matched WT animals, whereas Kl/Kl mice fed with a vitamin D–replete diet displayed, as expected, the typical hallmarks of Klotho deficiency (Fig. 5A, 5B). The thymus architecture of 8-wk-old, conventionally fed Kl/Kl mice showed a structural disorganization, especially of the medulla (Fig. 5C), consistent with our previous observations (47). In contrast, the histological structure of the thymus in both Kl/Kl and WT mice fed with a vitamin D–deprived diet displayed separate and well-demarcated cortical and medullary compartments (Fig. 5C). These data specify that low vitamin D levels correct the thymus phenotype of Kl/Kl mice and suggest that the premature thymus involution observed in Kl/Kl mice is the result of high systemic levels of vitamin D.

We considered the possibility that hypervitaminosis D in Kl/Kl mice may inhibit the emigration of mature thymocytes to the periphery, possibly contributing to T cell proliferation and/or survival, thereby contributing to T cell lymphopenia. We therefore probed in Kl/Kl and control mice the cellularity and frequency of CD4 and CD8 SP T (CD4 T and CD8 T) cells in spleen and lymph nodes. At 8 wk of age, Kl/Kl mice fed a vitamin D–depleted diet had comparable percentages (Fig. 6A) but decreased numbers (Fig. 6B) of both CD4 T and CD8 T cells compared with controls.

An increased frequency of CD44^hi memory-like cells is a phenotypic hallmark of T cell lymphopenia. Interestingly, Kl/Kl mice also may have impaired IL-7–dependent lymphopenia-driven differentiation, expansion, or survival of memory-like T cells because of decreased IL-7 production in stromal cells (e.g., BM) in Kl/Kl mice (48). The frequency of CD4 T CD44^hi but not CD8 T CD44^hi memory-like T cells was unaffected (Fig. 6C, Supplemental Fig. 2), although the cellularity of both was reduced in Kl/Kl mice on a vitamin D–depleted diet (Fig. 6D), suggesting a defect in memory-like T cell support. Moreover, the absolute cellularity of splenic CD4 T and CD8 T cell lymphopenia was comparable between Kl/Kl mice fed a vitamin D–deprived diet and control animals (Fig. 6E, 6F). Overall, the data show that Kl/Kl animals at 8 wk of age have significant losses of both thymocytes and TECs and that these deficits can be overcome by decreasing vitamin D levels in vivo.

Our results demonstrate that thymus function in young Kl/Kl mice is unaffected, but severe atrophy with reduced thymopoiesis is observed in these mice by 8 wk of age. This change in thymus function was neither the consequence of a cell-intrinsic loss of Klotho expression in hematopoietic cells nor the result of a cell-autonomous deficiency in expression of this protein by thymic stromal cells. Rather, thymic hypocellularity and the loss of a regular thymic stromal architecture in Kl/Kl mice are the consequence of high vitamin D levels, suggesting that Klotho deficiency disrupts thymus function in an indirect, systemic fashion.

The molecular interactions of FGF23, Klotho, and vitamin D coordinate to regulate phosphate metabolism (24, 46, 49–51). FGF23 decreases renal tubular and phosphate reabsorption and stimulates vitamin D₃, which results in increased renal Klotho synthesis (46). Klotho binds to FGFR1(IIIc), which reduces vitamin D₃ synthesis. In Kl/Kl mice, FGF23 is unable to bind the Klotho transmembrane molecule, precluding its transport to FGFR1c, resulting in a failure to negatively regulate vitamin D₃ synthesis via FGF23 receptor-mediated inhibition of 1α-hydroxylase (52).Thus, high levels of active vitamin D accumulate in Kl/Kl mice, causing a state of calcium and phosphate imbalance.

To counteract the abnormal calcium and phosphorus state in kl/kl mice, vitamin D–deficient diets have been tested and have been shown to ameliorate other consequences related to accelerated aging, including decreased ectopic calcification of tissues, lack of skin atrophy, and increased life span (45). Feeding kl/kl mice a vitamin D–deficient diet indeed improved thymic architecture, TEC differentiation, thymocyte development and maturation, and peripheral T cell reconstitution. Therefore, it may be tempting to speculate that the vitamin D–deficient diet had direct effects on TECs, despite the fact that the genetic deficiency of Klotho in TECs would be unaffected. Similarly, the data from kidney capsule implants indicated that the TECs from kl/kl mice are able to develop and support thymopoiesis despite the Klotho deficiency. Taken together, we favor the hypothesis that correction of calcium and phosphorus imbalance and high vitamin D levels are responsible for improvement in thymopoiesis and TEC development.

Both indirect and direct TEC effects of Klotho deficiency can be envisioned. Metabolic imbalance of calcium, phosphorus, and vitamin D₃ may indirectly impair TEC function and development because of apoptosis and stress in the animals.

High vitamin D₃ levels may indirectly impair TEC development and maturation by interfering with the essential cross-talk signaling mechanisms between TECs and thymocytes needed for their mutual development (43, 53). In support of this hypothesis, thymocytes express the vitamin D receptor according to gene expression data (accession no. GSE81163, http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE81163). High vitamin D levels directly impair TEC development because cells within the thymus that express high levels of vitamin D receptor are sensitive to vitamin D–induced signaling. In the absence of Klotho, levels of FGF23 are elevated (54), which may be toxic to developing thymocytes, as it is to the kidney (55), or exert yet unknown effects independent of vitamin D levels. Klotho suppresses NF-κB translocation and therefore suppresses inflammatory cytokine production (56), and conversely, Klotho deficiency increases proinflammatory cytokines such as IL-6 or TNF-α in monocytes (57), which could be driving thymocyte death, leading to insufficient support of TECs.

Medullary thymocytes also express the vitamin D receptor, and vitamin D signaling inhibits mitogenic stimulation of these cells (58, 59), which may negatively impact peripheral T lineage reconstitution. We observed a reduction of Tregs in the thymus and in the periphery of 8-wk-old Kl/Kl mice (data not shown). Thymic Tregs express vitamin D receptor, suggesting that their development could be affected by vitamin D levels in a manner similar to conventional thymocytes (60). Additional potential mechanisms that might lead to thymic involution and T cell lymphopenia seen in Kl/Kl mice include decreased expression of positive cell cycle regulators, such as cyclin D1 and c-Myc, which may be lacking in developing thymocytes, resulting in low proliferative rates. Previously, we reported that keratinocyte growth factor (fibroblast growth factor-7) administration, known to stimulate the proliferation of TECs and other epithelial cells, can improve IL-7 production and thymopoiesis in 2-wk- but not 6-wk-old Kl/Kl mice that have substantially defective IL-7 production and thymopoiesis (47). Although it is not yet clear whether reduction of IL-7 production directly associates with hypervitaminosis D, mice given a vitamin D–deficient diet did not have evidence of poor T cell content in the periphery or manifestations of IL-7 deficiency in the thymus. Further studies are needed to identify the molecular mechanisms downstream of reduced vitamin D levels that account for the preservation of regular thymopoiesis in Kl/Kl mice.

In conclusion, our data suggest that Klotho is not cell-autonomously needed for differentiation, maintenance, or function of the thymic microenvironment. However, under conditions that stress the thymus and developing T cells, Klotho is required to maintain normal levels of thymopoiesis. In this way, Klotho may be important in opposing the process of thymic involution, which naturally occurs with age.

The authors have no financial conflicts of interest.